add some comments

python/paddle/v2/data_feeder.py

@@ -15,5 +15,46 @@
 from py_paddle import DataProviderConverter
 
 __all__ = ['DataFeeder']
+"""
+DataFeeder converts the data returned by paddle.reader into a data structure
+of Arguments which is defined in the API. The paddle.reader usually returns
+a list of mini-batch data. Each item in the list is a tuple or list, which is
+one sample with multiple features. DataFeeder converts this mini-batch data
+into Arguments in order to feed it to C++ interface.
+
+The example usage:
+
+    data_types = [paddle.data_type.dense_vector(784),
+                  paddle.data_type.integer_value(10)]
+    feeder = DataFeeder(input_types=data_types)
+    minibatch_data = [
+                       ( [1.0,2.0,3.0,4.0], 5, [6,7,8] ),  # first sample
+                       ( [1.0,2.0,3.0,4.0], 5, [6,7,8] )   # second sample
+                     ]
+
+    #  or 
+    #  minibatch_data = [
+    #                     [ [1.0,2.0,3.0,4.0], 5, [6,7,8] ],  # first sample
+    #                     [ [1.0,2.0,3.0,4.0], 5, [6,7,8] ]   # second sample
+    #                   ]
+    arg = feeder(minibatch_data)
+
+
+Args:
+    input_types: A list of input data types. It's length is equal to the length
+                 of data returned by paddle.reader. Each item specifies the type
+                 of each feature.
+    mintbatch_data: A list of mini-batch data. Each item is a list or tuple,
+                    for example:
+                    [ 
+                      (feature_0, feature_1, feature_2, ...), # first sample
+                      (feature_0, feature_1, feature_2, ...), # second sample
+                      ...
+                    ]
+
+Returns:
+    An Arguments object contains this mini-batch data with multiple features.
+    The Arguments definition is in the API.
+"""
 
 DataFeeder = DataProviderConverter

python/paddle/v2/trainer.py

@@ -89,7 +89,6 @@ class SGD(ITrainer):
             event_handler = default_event_handler
 
         topology = v2_layer.parse_network(topology)
-        print topology
 
         __check_train_args__(**locals())